US3803395A - Method of an apparatus for processing of signals to transform information between the time domain and frequency domain - Google Patents

Method of an apparatus for processing of signals to transform information between the time domain and frequency domain Download PDF

Info

Publication number
US3803395A
US3803395A US00286929A US28692972A US3803395A US 3803395 A US3803395 A US 3803395A US 00286929 A US00286929 A US 00286929A US 28692972 A US28692972 A US 28692972A US 3803395 A US3803395 A US 3803395A
Authority
US
United States
Prior art keywords
pulses
acoustic
predetermined
group
signal
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Lifetime
Application number
US00286929A
Inventor
C Quate
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Leland Stanford Junior University
Original Assignee
Leland Stanford Junior University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Leland Stanford Junior University filed Critical Leland Stanford Junior University
Priority to US00286929A priority Critical patent/US3803395A/en
Application granted granted Critical
Publication of US3803395A publication Critical patent/US3803395A/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06GANALOGUE COMPUTERS
    • G06G7/00Devices in which the computing operation is performed by varying electric or magnetic quantities
    • G06G7/12Arrangements for performing computing operations, e.g. operational amplifiers
    • G06G7/19Arrangements for performing computing operations, e.g. operational amplifiers for forming integrals of products, e.g. Fourier integrals, Laplace integrals, correlation integrals; for analysis or synthesis of functions using orthogonal functions
    • G06G7/195Arrangements for performing computing operations, e.g. operational amplifiers for forming integrals of products, e.g. Fourier integrals, Laplace integrals, correlation integrals; for analysis or synthesis of functions using orthogonal functions using electro- acoustic elements

Definitions

  • ABSTRACT A method of and apparatus for signal processing to transform information between the time domain and frequency domain (Fourier transform) wherein a time-varying electromagnetic signal to be analyzed and a series of radio frequency pulses are applied to a piezoelectric medium so as to interact at predetermined positions, and generate sampling pulses which are subsequently combined in the medium in a fashion to provide the discrete Fourier transform of the timevarying signal.
  • the present invention relates generally to the processing of electromagnetic signals and more particularly to a method of and apparatus for processing electromagnetic signals in a fashion to provide transformation of information between a-time domain and a frequency domain such as the Fourier, Hadamard, or analogous transforms.
  • the Fourier transform has for many years been used for identifying the frequency components making up a continuouswaveform of an electromagnetic signal or, in other words, to determine the frequency spectrum of asignal which is presented as a running variation with time.
  • the continuous Fourier transform F(w), or spectrum, of a time-varyingelectromagnetic signal, f(t) is given by
  • an excellent approximation of the continuous Fourier transform has been provided by the socalled discrete Fourier transform (DFT) wherein the time-varying signal to be analyzed is sampled at discrete time intervals.
  • DFT discrete Fourier transform
  • a derivation of the discrete Fourier transform from the continuous Fourier transform can be found for example in the volume by Blackman and Tukey entitled, The Measurement of Power Spectra, New York; Dover, 1958, and can be represented by the following summation where N is the total number ofsamples in the time domain,
  • n the numbering of the samples in the time domain
  • m represents the numbering of the discrete samples in the frequency domain
  • W is the complex exponential, exp (-721r/N)
  • A(n) are the time domainsample values.
  • the discrete Fourier transform lends itself to digital pro cessing' techniques which are discussed theoretically in such article and whcih have been implemented interms of hardward in the entire issue of IEEE Transactions on Audio and Electroacoustics I such as the Hadamard transform discussed in detail in the article by Pratt, Kane and Andrews entitled Hadamard Transform image Coding, IEEE Proceedings, Vol. 57, No. l, p. 58 ff. (January 1969), and the same processing problems have been encountered.
  • the method is carried out through the interaction of acoustic waves in a piezoelectric medium.
  • the signal wave to be analyzed which takes the form of a. time-varying electromagnetic signal is introduced by a suitable directional transducer of known characteristics to the piezoelectric medium so as to traverse the medium as a continuous acoustic wave which can be of any type but is preferably an acoustic surface wave.
  • the acoustic signal wave is sampled at discrete intervals by the introduction of a series of identical short acoustic pulses by a suitable directional transducer so as to travel along the same predetermined path in a fashion such that translation of the pulses with respect to the acoustic signal wave occurs.
  • a standard pulse generator is initially utilized to form the series of identical electromagnetic pulses, preferably at regular intervals and of predetermined phase at a frequency which is not critical but preferably is at a radio frequency of approximately I00 MHz.
  • the individual pulses are identical, .each having the same amplitude and a pulse duration of preferably 50 nanoseconds so that when applied .to the piezoelectric medium they will occupy a distance along the predetermined propagation path of approximately microns.
  • the number of radio frequency sampling pulses in a given period of time is selected so that a plurality of the sampling pulses will occur during a single cycle of the signal wave which is to be analyzed, and a corresponding number of output transducers are disposed at regularly spaced positions along the common predetermined propagation path of the signal wave and the sampling acoustic pulses enabling the extraction of electromagnetic energy at each transducer when an individual pulse and one portion of the acoustic signal wave overlap or obtain phase'coherence under a particular transducer, thus to provide nonlinear parametric interaction in a known fashion so to generate within the piezoelectric medium as an electric polarization which is extracted as electrogmagnetic energy by the output transducer.
  • the output at a given transducer is proportional to the product of the instantaneous value of the signal wave, and the sampling pulse, but in view of the fact that all sampling pulses are identical, it will be clear that the output at each individual transducer will be directly proportional to the amplitude of the signal wave at that position.
  • the electromagnetic energy extracted at each output transducer position resultant from the parametric interaction of the signal wave and one of the sampling pulses is in turn used to excite a unique one of a first predetermined set of electro-acoustic transducers, thus to generate at unique positions in the piezoelectric medium a group of acoustic pulses along predetermined paths.
  • pulses extracted by the first receiving transducer are then used to excite one of a second set of electroacoustic transducers appropriately, positioned on the same piezoelectric medium to in turn generate a second group of acoustic signals along predetermined paths for additionalsummation at second receiving transducers, suchrepetitive summation or combining process continuing to provide the series summation generally expressed by the previously mentioned mathematical series summation:
  • DFT 2 A(mW pling pulses at predetermined and preferably regular time intervals, and applying those sampling pulses to the piezoelectric medium to traverse the same path in a fashion such that translation of the sampling pulses occurs relative to the time-varying signal and parametric coupling of the individual pulses and the signal wave occurs at predetermined spaced positions whereat the coupled energy may be extracted by suitable transducers-to generate acoustic pulses in.a fashion so that repeated summations of acoustic energy corresponding to the mathematical formulation of the discrete Fourier transform (DFT) or other transforms is attained.
  • DFT discrete Fourier transform
  • FIG 1 is'a diagrammatic representation of a specific acoustic wave device-arranged to perform a fastdiscrete F ourier'transform at eight sample positions of one cycle of the signal wave
  • FIG, 2 is a graphical representation of a signal'wave in its time and frequency domains.
  • v a signal wave flr) as illustrated in the upperpbrtionlof FIG. 2 extending for a cyclical period of 2T in the time domain is sampled at intervals equal to B with n representing the discrete numbering of the samples in such time domain, the total samples equalling N (2T)/(% lnth'e frequency domain, shown in the bottom poriiong f 116.
  • the to tal frequency spectrum is indicated as extending over a bandwidth of ffi a nd the frequency samples are in turn'spaced by an interval of 95 T, the 1 discrete numbering of the samples in the frequency domain beingindicated by m.
  • the signal wave f(t) is delivered through an input directional electro-acoustic transducer which preferably takes the form of an interdigital structure such as described in the article Convo- -lution and Time Inversion Using Parametric Interactions of Acoustic Surface Waves by Luukkala and Kino in Applied Physics Letters, volume 18, Number 9, May I, 1971, so as to generate in a piezoelectric medium 12 composed of lithium niobate, bismuth germa eight of the sampling pulses occupy the same time pe riod as a single cycle of the signal wave.
  • the series of pulses are delivered to a directional electro-acoustic transducer 16 also of the same interdigital type to the left end of the piezoelectric medium 12 so as to generate a series of acoustic pulses which travel from left to right along the same propagation path as the signal wave but in the opposite direction so that translation of the two acoustic waves occurs.
  • the two signals that is, the series of sampling pulses and the signal wave, are caused to parametrically interact in a fashion briefly mentioned in the referred to Luukkala-Kino article and explained in detail in Chapter 5 of W. H. Louisell, Coupled Mode and Parametric Electronics, 1960. Briefly, if two signals of frequency w, and m and propagation constants k and k respectively are passed in opposite directions through a piezoelectric medium, the nonlinearity present in the medium gives rise to an acoustic strain component with a frequency 0);, m, m and a propagation constant k k,k It may be mentioned that it is now well known that if the two waves are surface acoustic waves,
  • the power densities arehigh so that the acoustic strain is relatively large whereby the nonlinear parametric interaction can take place at relatively low power densities.
  • the signals can be controlled to provide a combined propagation constant k,, 0 in which case, as explained in the mentioned Luukkala- Kino article, electromagnetic energy proportional to the product of the strain amplitudes of an individual pulse and the overlapping portion of the signal wave can be extracted by a sampling transducer 18 in the form of a gold electrode pair or by a semiconductor arrangement as described in the Lee-Gunshor article, Enhancement of Nonlinearity in Surface Acoustic Wave Propagation from Coupling to Charge Carriers",
  • the output electromagnetic energy is proportional to the level of the sampled signal wave, it being remembered that the sampling pulses are identical and thus give rise to equivalent strain amplitudes in a piezoelectric medium.
  • the outputs of the eight sampling transducers l8 constitute the eight samples of the signal wave designated as A(O), A(l), A(2), A(3), A(4), A(S), A(6) and A(7).
  • the electromagnetic outputs of the eight sampling transducers are delivered by suitable electric connections to spaced sets of electroacoustic transducers Applied Physics Letters, volume20, Number 8, April 22, preferably in the form of interdigital transducers, which generate acoustic propagating waves along a group of parallel paths, as diagrammatically indicated, selected pairs of the transducers being longitudinally aligned along common parallel acoustic paths which extends to corresponding sum transducers 24 which convert the sum of the acoustic wave pairs along each common propagation path to an electromagnetic signal so as to correspond to the first summation section of the Table producing the A series of signals.
  • the sampled pulses A(O) and A(4) excite two transducers 22 which are in alignment, and the summation of these two generated acoustic signals is delivered to the uppermost sum transducer 24 to provide the summed signal A (0).
  • the respective transducers providing for this first summation series are appropriately spaced on the piezoelectric medium to provide signal overlay at the sum transducer and a phase change which is the acoustic analog of the multiplication by the complex exponential, W, it being well-known that a change in acoustic path length produces a phase change of the propagated signal.
  • the A, first sum transducers 24 in turn excite pairs of transducers 26 corresponding to the additive elements in the second Table section so as to generate acoustic pulses traveling in summed parallelism at right angles to the first summation of signals and are in turn extracted as electromagnetic signals by aligned second sum transducers 28 in the A series, and repetitively, these A; transducers excite additional interdigital transducers 30 that provide the final signal summation corresponding to the third additive series providing the A series output at the third sum transducers 32.
  • the sampled frequency components, X,, are directly related to the A series in the following fashion:
  • sampling pulse generating step includes generation .of a plurality of sampling pulses during one cycle of the time-varying electromagnetic signal.
  • the group of predetermined paths of the generated acoustic pulses are parallel and are nonintersecting with other paths of acoustic waves.
  • radio frequency pulse generator for producing a series of short, regular, identical pulses
  • transducer means for applying the short pulses to said mediumto generate acoustic pulses along a predetermined path

Landscapes

  • Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Acoustics & Sound (AREA)
  • Software Systems (AREA)
  • Computer Hardware Design (AREA)
  • General Physics & Mathematics (AREA)
  • Surface Acoustic Wave Elements And Circuit Networks Thereof (AREA)

Abstract

A method of and apparatus for signal processing to transform information between the time domain and frequency domain (Fourier transform) wherein a time-varying electromagnetic signal to be analyzed and a series of radio frequency pulses are applied to a piezoelectric medium so as to interact at predetermined positions, and generate sampling pulses which are subsequently combined in the medium in a fashion to provide the discrete Fourier transform of the time-varying signal.

Description

United States Patent [1 1 Quate 1451 Apr. 9, 1974 METHOD OF AN APPARATUS FOR PROCESSING OF SIGNALS TO TRANSFORM INFORMATION BETWEEN THE TIME DOMAIN AND FREQUENCY DOMAIN [75] Inventor: Calvin F. Quate, Los Altos Hills,
Calif.
[73 Assignee: The Board of Trustees of Leland Stanford Junior University, Stanford, Calif.
22 Filed: Sept. 7, 1972 211 Appl. No.: 286,929
[52] US. Cl. 235/193, 235/197, 3l0/8.l [5|] Int.'CI. G06g 7/19, HOlv 7/00 [58] Field of Search 235/181, 183, 193, 194, 235/197; 444/1; 310/8.l
[56] References Cited UNITED STATES PATENTS 3,675,052 7/1972 Lindsay et al. 3 l0/8.l
3/1972 Lean et a1 3l0/8.l
OTHER PUBLICATIONS Fisk: New Acoustoelectronic Compounds Replace Conventional Tuned Circuits, Popular Electronics, March 1971, pp. 27-33.
Primary Examiner-Felix D. Gruber Attorney, Agent, or Firm-Paul B. Fihe [5 7 ABSTRACT A method of and apparatus for signal processing to transform information between the time domain and frequency domain (Fourier transform) wherein a time-varying electromagnetic signal to be analyzed and a series of radio frequency pulses are applied to a piezoelectric medium so as to interact at predetermined positions, and generate sampling pulses which are subsequently combined in the medium in a fashion to provide the discrete Fourier transform of the timevarying signal. a
5 Claims, 2 Drawing Figures RF PULSE 6ENERATOR -mPur SIGNAL H1) A lo) A10) A m la lm lm rsu is) A 11) 28 j A 10) W 4 A3(\) I 1 W H I 30 A (z) Ill WW METHOD OF AN A PARATus FOR PROCESSING OF SIGNALS TO TRANSFORM INFORMATION BETWEEN THE TIM DOMAIN AND FREQUENCY D MAIN FIELD OF THE INVENTION The present invention relates generally to the processing of electromagnetic signals and more particularly to a method of and apparatus for processing electromagnetic signals in a fashion to provide transformation of information between a-time domain and a frequency domain such as the Fourier, Hadamard, or analogous transforms.
BACKGROUND OFTHE INVENTION The Fourier transform has for many years been used for identifying the frequency components making up a continuouswaveform of an electromagnetic signal or, in other words, to determine the frequency spectrum of asignal which is presented as a running variation with time. Mathematically, the continuous Fourier transform F(w), or spectrum, of a time-varyingelectromagnetic signal, f(t) is given by In practice, an excellent approximation of the continuous Fourier transform has been provided by the socalled discrete Fourier transform (DFT) wherein the time-varying signal to be analyzed is sampled at discrete time intervals. A derivation of the discrete Fourier transform from the continuous Fourier transform can be found for example in the volume by Blackman and Tukey entitled, The Measurement of Power Spectra, New York; Dover, 1958, and can be represented by the following summation where N is the total number ofsamples in the time domain,
n represents the numbering of the samples in the time domain,
m represents the numbering of the discrete samples in the frequency domain W is the complex exponential, exp (-721r/N), and
A(n) are the time domainsample values. As explained in an article entitled, A Guided Tour of the Fast Fourier Transform by G. D. Bergland in IEEE Spectrum of July, 1969, pgs 41 ff, the discrete Fourier transform lends itself to digital pro cessing' techniques which are discussed theoretically in such article and whcih have been implemented interms of hardward in the entire issue of IEEE Transactions on Audio and Electroacoustics I such as the Hadamard transform discussed in detail in the article by Pratt, Kane and Andrews entitled Hadamard Transform image Coding, IEEE Proceedings, Vol. 57, No. l, p. 58 ff. (January 1969), and the same processing problems have been encountered.
SUMMARY OF THE PRESENT INVENTION lizing acoustic waves to process signals whereby information can be transformed between the time domain and the frequency domain in a relatively simple fashion with a minimum of both time and cost. Generally, the method is carried out through the interaction of acoustic waves in a piezoelectric medium. The signal wave to be analyzed which takes the form of a. time-varying electromagnetic signal is introduced by a suitable directional transducer of known characteristics to the piezoelectric medium so as to traverse the medium as a continuous acoustic wave which can be of any type but is preferably an acoustic surface wave. In the'piezoelectric'medium, the acoustic signal wave is sampled at discrete intervals by the introduction of a series of identical short acoustic pulses by a suitable directional transducer so as to travel along the same predetermined path in a fashion such that translation of the pulses with respect to the acoustic signal wave occurs. A standard pulse generator is initially utilized to form the series of identical electromagnetic pulses, preferably at regular intervals and of predetermined phase at a frequency which is not critical but preferably is at a radio frequency of approximately I00 MHz. As mentioned, the individual pulses are identical, .each having the same amplitude and a pulse duration of preferably 50 nanoseconds so that when applied .to the piezoelectric medium they will occupy a distance along the predetermined propagation path of approximately microns. The number of radio frequency sampling pulses in a given period of time is selected so that a plurality of the sampling pulses will occur during a single cycle of the signal wave which is to be analyzed, and a corresponding number of output transducers are disposed at regularly spaced positions along the common predetermined propagation path of the signal wave and the sampling acoustic pulses enabling the extraction of electromagnetic energy at each transducer when an individual pulse and one portion of the acoustic signal wave overlap or obtain phase'coherence under a particular transducer, thus to provide nonlinear parametric interaction in a known fashion so to generate within the piezoelectric medium as an electric polarization which is extracted as electrogmagnetic energy by the output transducer. The output at a given transducer is proportional to the product of the instantaneous value of the signal wave, and the sampling pulse, but in view of the fact that all sampling pulses are identical, it will be clear that the output at each individual transducer will be directly proportional to the amplitude of the signal wave at that position. g
The electromagnetic energy extracted at each output transducer position resultant from the parametric interaction of the signal wave and one of the sampling pulses is in turn used to excite a unique one of a first predetermined set of electro-acoustic transducers, thus to generate at unique positions in the piezoelectric medium a group of acoustic pulses along predetermined paths.
pulses extracted by the first receiving transducer are then used to excite one of a second set of electroacoustic transducers appropriately, positioned on the same piezoelectric medium to in turn generate a second group of acoustic signals along predetermined paths for additionalsummation at second receiving transducers, suchrepetitive summation or combining process continuing to provide the series summation generally expressed by the previously mentioned mathematical series summation:
DFT= 2 A(mW pling pulses at predetermined and preferably regular time intervals, and applying those sampling pulses to the piezoelectric medium to traverse the same path in a fashion such that translation of the sampling pulses occurs relative to the time-varying signal and parametric coupling of the individual pulses and the signal wave occurs at predetermined spaced positions whereat the coupled energy may be extracted by suitable transducers-to generate acoustic pulses in.a fashion so that repeated summations of acoustic energy corresponding to the mathematical formulation of the discrete Fourier transform (DFT) or other transforms is attained.
BRlEF DESCRlPTlON OF THE DRAWINGS The stated objective of the invention and the manner in which it is achieved as summarized hereinabove will be more readily understood by-reference to. the following detailed description of a specific embodiment of the invention as shown in the accompanying drawings:
' FIG 1 is'a diagrammatic representation of a specific acoustic wave device-arranged to perform a fastdiscrete F ourier'transform at eight sample positions of one cycle of the signal wave,- and FIG, 2 is a graphical representation of a signal'wave in its time and frequency domains.
DETAILED DESCRIPTION (ism EXE-MPLARY EM BODIMEN T OF THE INVENTION If one wishes to perform a discrete Fourier transform I in accordance with the mentione d summation,
v a signal wave flr) as illustrated in the upperpbrtionlof FIG. 2 extending for a cyclical period of 2T in the time domain is sampled at intervals equal to B with n representing the discrete numbering of the samples in such time domain, the total samples equalling N (2T)/(% lnth'e frequency domain, shown in the bottom poriiong f 116. g the to tal frequency spectrum is indicated as extending over a bandwidth of ffi a nd the frequency samples are in turn'spaced by an interval of 95 T, the 1 discrete numbering of the samples in the frequency domain beingindicated by m. i
The precise derivation of this series summation to compute DFT is explained in detail in the mentioned article, A Guided Tour of the Fast Fourier Transfonn by G. D. Bergland in the IEEE Spectrum of July, 1969, pgs 41 ff, to which reference is made for mathematical details. If one desires to perform the discrete Fourier transform with aisampling of the signal wave at eight regular intervals (N 8), the eight samples in the time domain can be designated by the symbols A(O), A(l A(2), A(3), A(4), M5), M6) and A (7), and following the Bergland calculation of the discrete Fourier transform, the initial eight samplesare used to compute three successive summations, as listed below, to enable ultimate calculation of the discrete frequency components: i l l A,(0) A(O) A(4)W A (1)= A(l) A(5,)W
3( z( 2( i 7 It will be seen that the individual mathematical operations are quite simple constituting a simpleadditive process coupled with multiplication by acomplex exponential. The addition process can be achieved quite simply in acoustic wave devices operating in a known fashion by the summation of two acoustic signals in an individual output transducer and, in turn, the acoustic analog of the multiplication by the complex exponential can, readily be achieved by a simple phase shift. With such general acoustic wave concepts in mind, the transformation of eight samples of a signalwave from the time domain into the frequency domain can readily be achieved by the apparatus diagrammatically indicated in FIG. I.
More particularly, the signal wave f(t) is delivered through an input directional electro-acoustic transducer which preferably takes the form of an interdigital structure such as described in the article Convo- -lution and Time Inversion Using Parametric Interactions of Acoustic Surface Waves by Luukkala and Kino in Applied Physics Letters, volume 18, Number 9, May I, 1971, so as to generate in a piezoelectric medium 12 composed of lithium niobate, bismuth germa eight of the sampling pulses occupy the same time pe riod as a single cycle of the signal wave. The series of pulses are delivered to a directional electro-acoustic transducer 16 also of the same interdigital type to the left end of the piezoelectric medium 12 so as to generate a series of acoustic pulses which travel from left to right along the same propagation path as the signal wave but in the opposite direction so that translation of the two acoustic waves occurs.
The two signals, that is, the series of sampling pulses and the signal wave, are caused to parametrically interact in a fashion briefly mentioned in the referred to Luukkala-Kino article and explained in detail in Chapter 5 of W. H. Louisell, Coupled Mode and Parametric Electronics, 1960. Briefly, if two signals of frequency w, and m and propagation constants k and k respectively are passed in opposite directions through a piezoelectric medium, the nonlinearity present in the medium gives rise to an acoustic strain component with a frequency 0);, m, m and a propagation constant k k,k It may be mentioned that it is now well known that if the two waves are surface acoustic waves,
the power densities arehigh so that the acoustic strain is relatively large whereby the nonlinear parametric interaction can take place at relatively low power densities.
While not essential, the signals can be controlled to provide a combined propagation constant k,, 0 in which case, as explained in the mentioned Luukkala- Kino article, electromagnetic energy proportional to the product of the strain amplitudes of an individual pulse and the overlapping portion of the signal wave can be extracted by a sampling transducer 18 in the form of a gold electrode pair or by a semiconductor arrangement as described in the Lee-Gunshor article, Enhancement of Nonlinearity in Surface Acoustic Wave Propagation from Coupling to Charge Carriers",
electric medium 12, and the spacing between the transduccrs is equivalent to the spacing between the sampling p'ulses along the propagation path. Accordingly, at each sampling transducer 18 the output electromagnetic energy is proportional to the level of the sampled signal wave, it being remembered that the sampling pulses are identical and thus give rise to equivalent strain amplitudes in a piezoelectric medium. Thus with reference to the previous mathematical representations, and as specifically indicated in the drawing, the outputs of the eight sampling transducers l8 constitute the eight samples of the signal wave designated as A(O), A(l), A(2), A(3), A(4), A(S), A(6) and A(7).
The electromagnetic outputs of the eight sampling transducers are delivered by suitable electric connections to spaced sets of electroacoustic transducers Applied Physics Letters, volume20, Number 8, April 22, preferably in the form of interdigital transducers, which generate acoustic propagating waves along a group of parallel paths, as diagrammatically indicated, selected pairs of the transducers being longitudinally aligned along common parallel acoustic paths which extends to corresponding sum transducers 24 which convert the sum of the acoustic wave pairs along each common propagation path to an electromagnetic signal so as to correspond to the first summation section of the Table producing the A series of signals. For example, the sampled pulses A(O) and A(4) excite two transducers 22 which are in alignment, and the summation of these two generated acoustic signals is delivered to the uppermost sum transducer 24 to provide the summed signal A (0). The respective transducers providing for this first summation series are appropriately spaced on the piezoelectric medium to provide signal overlay at the sum transducer and a phase change which is the acoustic analog of the multiplication by the complex exponential, W, it being well-known that a change in acoustic path length produces a phase change of the propagated signal.
The A, first sum transducers 24 in turn excite pairs of transducers 26 corresponding to the additive elements in the second Table section so as to generate acoustic pulses traveling in summed parallelism at right angles to the first summation of signals and are in turn extracted as electromagnetic signals by aligned second sum transducers 28 in the A series, and repetitively, these A; transducers excite additional interdigital transducers 30 that provide the final signal summation corresponding to the third additive series providing the A series output at the third sum transducers 32. The sampled frequency components, X,,, are directly related to the A series in the following fashion:
Many variations from the described exemplary apparatus can be envisioned. For example, there is no particular limit on the number of samples since acoustic devices with rather long delay paths (acoustic velocity 3 X 10 cm./sec.,) are known and power levels of acoustic surface waves are minimal. Furthermore, even though several acoustic summations are required, by
the arrangement of non-intersecting propagation paths, the complete transformation can be carried out on a single piezoelectric unit or, as-an alternative on several piezoelectric units which can be electrically connected. Additionally, the method and apparatus are obviously applicable to othertransforms such as the mentioned Hadamard transform. Accordingly, the foregoing description of one embodiment is not to be considered as limiting and the actual scope of the invention is to be indicated only by the appended claims.
What is claimed is: 1. The method of signal processing to transform information between' the time domain and the frequency domain which comprises the steps of extracting electromagnetic energy resultant from the parametric coupling of the sampling pulses and the signal wave at the predetermined spaced positions along the first predetermined path,
generating acoustic pulses at a first set of unique positions in the piezoelectric medium with the extracted electromagnetic energy corresponding to the coupled sampling pulses and signal wave to follow a first group of predetermined paths, at least two pulses being generated to follow each of said first group of predetermined paths, extracting electromagnetic energy at spaced positions resultant from the summation of acoustic pulses traversing each of the first group of predetermined paths, generating acoustic pulses at a second set of unique positions in the piezoelectric medium in response to the summation of the pulses extracted from each of the first group of predetermined paths, to follow a second group of predetermined paths, at least two pulses being generated to follow each of said second group of predetermined paths, extracting energy at spaced positions resultant from the summation of the acoustic pulses traversing each of the second group of predetermined paths,
generating acoustic pulses at a third set of unique positions in the piezoelectric medium in response to the summation of the pulses extracted from each of the second group of predetermined paths, to follow a third group of predetermined paths, at least two pulses being generated to follow each of said third groupof predetermined paths, and extracting energy at spaced positions resultant from the summation of the acoustic pulses traversing each of the third group of predetermined paths.
2. The method of signal processing according to claim 1 wherein said sampling pulse generating step includes generation .of a plurality of sampling pulses during one cycle of the time-varying electromagnetic signal.
3. The method of signal processing according to claim 1 wherein said time-varying electromagnetic signal and said sampling pulses are applied to the piezoelectric medium to traverse a common path but in opposite directions.
4. The method of signal processing according to claim 1 the group of predetermined paths of the generated acoustic pulses are parallel and are nonintersecting with other paths of acoustic waves.
5. Signal processing apparatus to provide a discrete Fourier transform which comprises,
a radio frequency pulse generator for producing a series of short, regular, identical pulses,
a piezoelectric medium,
transducer means for applying the short pulses to said mediumto generate acoustic pulses along a predetermined path,
means for applying a signal to be analyzed to said piezoelectric medium to generate an acoustic wave along the same path in the opposite. direction,
a series of sampling transducers spaced along the common propagation path to extract energy resultant from coupling of the pulses and the signal wave,
a set of unique generating transducer pairs connected to said sampling transducers to be excited thereby and to generate acoustic pulses along predetermined'paths, I I
a first set of sum transducers for receiving acoustic energy from unique pairs of said transducer pairs,

Claims (5)

1. The method of signal processing to transform information between the time domain and the frequency domain which comprises the steps of applying a time-varying electromagnetic signal to a piezoelectric medium to generate an acoustic wave along a first predetermined path, generating a series of short electromagnetic sampling pulses at predetermined time intervals, applying the sampling pulses to the piezoelectric medium to traverse the same first predetermined path in a fashion such that translation of the sampling pulses occurs relative to the time-varying signal and parametric coupling of individual sampling pulses and the signal wave occurs at predetermined spaced positions, extracting electromagnetic energy resultant from the parametric coupling of the sampling pulses and the signal wave at the predetermined spaced positions along the first predetermined path, generating acoustic pulses at a first set of unique positions in the piezoelectric medium with the extracted electromagnetic energy corresponding to the coupled sampling pulses and signal wave to follow a first group of predetermined paths, at least two pulses being generated to follow each of said first group of predetermined paths, extracting electromagnetic energy at spaced positions resultant from the summation of acoustic pulses traversing each of the first group of predetermined paths, generating acoustic pulses at a second set of unique positions in the piezoelectric medium in response to the summation of the pulses extracted from each of the first group of predetermined paths, to follow a second group of predetermined paths, at least two pulses being generated to follow each of said second group of predetermined paths, extracting energy at spaced positions resultant from the summation of the acoustic pulses traversing each of the second group of predetermined paths, generating acoustic pulses at a third set of unique positions in the piezoelectric medium in response to the summation of the pulSes extracted from each of the second group of predetermined paths, to follow a third group of predetermined paths, at least two pulses being generated to follow each of said third group of predetermined paths, and extracting energy at spaced positions resultant from the summation of the acoustic pulses traversing each of the third group of predetermined paths.
2. The method of signal processing according to claim 1 wherein said sampling pulse generating step includes generation of a plurality of sampling pulses during one cycle of the time-varying electromagnetic signal.
3. The method of signal processing according to claim 1 wherein said time-varying electromagnetic signal and said sampling pulses are applied to the piezoelectric medium to traverse a common path but in opposite directions.
4. The method of signal processing according to claim 1 the group of predetermined paths of the generated acoustic pulses are parallel and are non-intersecting with other paths of acoustic waves.
5. Signal processing apparatus to provide a discrete Fourier transform which comprises, a radio frequency pulse generator for producing a series of short, regular, identical pulses, a piezoelectric medium, transducer means for applying the short pulses to said medium to generate acoustic pulses along a predetermined path, means for applying a signal to be analyzed to said piezoelectric medium to generate an acoustic wave along the same path in the opposite direction, a series of sampling transducers spaced along the common propagation path to extract energy resultant from coupling of the pulses and the signal wave, a set of unique generating transducer pairs connected to said sampling transducers to be excited thereby and to generate acoustic pulses along predetermined paths, a first set of sum transducers for receiving acoustic energy from unique pairs of said transducer pairs, a second set of pulse generating transducer pairs connected to said first set of sum transducers to generate a second set of acoustic pulses, a second set of sum transducers for receiving acoustic energy from unique pairs of said second pulse generating transducers, a third set of pulse generating transducer pairs connected to said second set of sum transducers to generate a third set of acoustic pulses, and a third set of sum transducers for receiving acoustic energy from said third set of pulse generating transducer pairs.
US00286929A 1972-09-07 1972-09-07 Method of an apparatus for processing of signals to transform information between the time domain and frequency domain Expired - Lifetime US3803395A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US00286929A US3803395A (en) 1972-09-07 1972-09-07 Method of an apparatus for processing of signals to transform information between the time domain and frequency domain

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US00286929A US3803395A (en) 1972-09-07 1972-09-07 Method of an apparatus for processing of signals to transform information between the time domain and frequency domain

Publications (1)

Publication Number Publication Date
US3803395A true US3803395A (en) 1974-04-09

Family

ID=23100755

Family Applications (1)

Application Number Title Priority Date Filing Date
US00286929A Expired - Lifetime US3803395A (en) 1972-09-07 1972-09-07 Method of an apparatus for processing of signals to transform information between the time domain and frequency domain

Country Status (1)

Country Link
US (1) US3803395A (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4049958A (en) * 1975-03-03 1977-09-20 Texas Instruments Incorporated Programable filter using chirp-Z transform
US4083021A (en) * 1975-10-17 1978-04-04 Thomson-Csf Transducer for surface wave filters with reduced diffraction
US4106103A (en) * 1976-07-19 1978-08-08 Xerox Corporation Derivation of discrete Fourier transform components of a time dependent signal
US4114116A (en) * 1977-08-01 1978-09-12 United Technologies Corporation Two-dimensional surface acoustic wave signal processor
US4191933A (en) * 1978-12-07 1980-03-04 United Technologies Corporation Differential mixing surface acoustic wave signal processor
US4321565A (en) * 1980-07-18 1982-03-23 Lockheed Missiles & Space Co., Inc. Relative phase sensor employing surface acoustic waves
US4385377A (en) * 1978-01-16 1983-05-24 Texas Instruments Incorporated Pulse position time division surface wave device transmitter
EP0136114A2 (en) * 1983-09-02 1985-04-03 Trw Inc. Signal processing system
US4764701A (en) * 1986-12-30 1988-08-16 Zenith Electronics Corporation Multichannel surface acoustic wave encoder/decoder
US4947073A (en) * 1988-02-04 1990-08-07 Trw Inc. Saw channelized filters
US4995006A (en) * 1988-03-31 1991-02-19 Wiltron Company Apparatus and method for low-pass equivalent processing

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4049958A (en) * 1975-03-03 1977-09-20 Texas Instruments Incorporated Programable filter using chirp-Z transform
US4083021A (en) * 1975-10-17 1978-04-04 Thomson-Csf Transducer for surface wave filters with reduced diffraction
US4106103A (en) * 1976-07-19 1978-08-08 Xerox Corporation Derivation of discrete Fourier transform components of a time dependent signal
US4114116A (en) * 1977-08-01 1978-09-12 United Technologies Corporation Two-dimensional surface acoustic wave signal processor
US4385377A (en) * 1978-01-16 1983-05-24 Texas Instruments Incorporated Pulse position time division surface wave device transmitter
US4191933A (en) * 1978-12-07 1980-03-04 United Technologies Corporation Differential mixing surface acoustic wave signal processor
US4321565A (en) * 1980-07-18 1982-03-23 Lockheed Missiles & Space Co., Inc. Relative phase sensor employing surface acoustic waves
EP0136114A2 (en) * 1983-09-02 1985-04-03 Trw Inc. Signal processing system
EP0136114A3 (en) * 1983-09-02 1987-09-02 Trw Inc. Signal processing system and method
US4764701A (en) * 1986-12-30 1988-08-16 Zenith Electronics Corporation Multichannel surface acoustic wave encoder/decoder
US4947073A (en) * 1988-02-04 1990-08-07 Trw Inc. Saw channelized filters
US4995006A (en) * 1988-03-31 1991-02-19 Wiltron Company Apparatus and method for low-pass equivalent processing

Similar Documents

Publication Publication Date Title
US3803395A (en) Method of an apparatus for processing of signals to transform information between the time domain and frequency domain
Joshi et al. Excitation and detection of surface elastic waves in piezoelectric crystals
CA1084590A (en) Selective intermodulation distortion measurement
Heffner Analysis of the backward-wave traveling-wave tube
Tseng Frequency response of an interdigital transducer for excitation of surface elastic waves
US3816753A (en) Parametric acoustic surface wave apparatus
US3803391A (en) A real-time processing system and method utilizing discrete fourier transform
US3760172A (en) Method of and apparatus for signal processing
Lerner The representation of signals
US3544894A (en) Apparatus for performing complex wave analysis
US3636333A (en) Fourier coefficient generator
US4051357A (en) Double odd discrete fourier transformer
US4016514A (en) Diode coupled tapped acoustic delay line correlator and convolver
Young et al. 'Complementary'signals and orthogonalized exponentials
US3925648A (en) Apparatus for the generation of a high capacity chirp-Z transform
US3543012A (en) Universal digital filter and function generator
Barker et al. System identification using pseudorandom signals and the discrete Fourier transform
US4296374A (en) Wideband digital spectrometer
GB1344084A (en) Noise-excited resonance apparatus
US2960664A (en) Wide band noise source
Bennett A general review of linear varying parameter and nonlinear circuit analysis
Bongianni Pulse compression using nonlinear interaction in a surface acoustic wave convolver
Roberts et al. Design and application of real-time spectrum-analyser systems
Richie et al. Description and development of a SAW filter CAD system
US3882408A (en) Parametric acoustic surface wave amplifier